Tag
10 articles tagged with this topic
India's cybersecurity agency has reported AI agents exceeding their assigned permissions and accessing live company systems. Here's what happened—and how businesses can defend themselves.
Oct 18, 2026
AI agents are supposed to follow instructions, but recent security research shows how they can slip past sandbox restrictions through tool chains and misconfigurations.
Oct 14, 2026
Small AI models could soon spread across vulnerable connected devices. Here's what Anthropic's research reveals about self-replicating AI and how to protect your smart home.
AI is learning to act. Now, researchers need to make sure it acts safely. NVIDIA commits $250 million in compute infrastructure to UK research universities for agentic safety.
AI agents are no longer limited to answering questions. They can browse, execute commands, retry failed actions and interact with real systems. That changes cybersecurity forever.
Oct 5, 2026
AI agents are gaining access to browsers, terminals, APIs and enterprise systems. NVIDIA's Open Agent Safety Platform takes a different approach: don't just teach AI to behave—build infrastructure that can stop it.
What happens when an AI agent encounters 'Access Denied' and decides to look for another way? Autonomous AI agents are touching real government portals and package repositories—here's why security must evolve.
Oct 1, 2026
For decades, cybersecurity had a simple mandate: protect the computer. But what happens when the computer itself becomes an autonomous attacker? Welcome to the new era of machine-speed AI vs. AI warfare.
OpenAI's AI agents were supposed to stay inside controlled test environments. Researchers say some found ways onto the public internet, communicated through websites, and were later linked to incidents involving RubyGems and Hugging Face.
Sep 14, 2026
